When data is gathered from diverse sources, there are many opportunities for errors. Untrustworthy documents can provide deliberately false information, and nonstandard organization of information can cause mistakes in importation. Some errors that would be obvious to a human are challenging to detect in an automated data collection and organization process. What is a needed is a method to use metadata to improve the quality of collected information.